FAQ
      Back to home
      1. Knowledge Base | OpsLogix
      2. VMware Management Pack
      3. FAQ

      Overcommitment Monitors in the VMware Management Pack

      This article describes which monitors are used in the VMware Management Pack and how they are calculated.

      The VMware Management Pack for SCOM includes a set of monitors to track overcommitment in your virtual environment. These monitors help you ensure that resources are being used efficiently and provide warnings and critical alerts when overcommitment exceeds defined thresholds.

      There are three sets of three monitors, categorized as follows:

      • All
      • Powered Off VMs
      • Powered On VMs

      Each category contains three types of monitors:

      • CPU
      • Datastore
      • Memory

      Thresholds

      Each monitor has specific thresholds to generate alerts:

      • Warning: Commitment percentage above 90%
      • Critical: Commitment percentage above 120%

      Calculation Methods

      The overcommitment metrics for each type of monitor are calculated using the following methods:

      Memory Overcommitment

      1. All Memory:

        hostMemoryPressure = 100 + ((totalUsedMemory - hostMemory) / hostMemory * 100)

      2. Powered Off VMs Memory:

        hostMemoryPressurePoweredOff = 100 + ((totalUsedMemoryPoweredOff - hostMemory) / hostMemory * 100)

      3. Powered On VMs Memory:

        hostMemoryPressurePoweredOn = 100 + ((totalUsedMemoryPoweredOn - hostMemory) / hostMemory * 100)

      Datastore Overcommitment

      1. All Datastore:

        DatastoreStoragePressureAll = 100 + ((allVmProvisioned - capacity) / capacity * 100)

      2. Powered On VMs Datastore:

        datastoreStoragePressurePoweredOn = 100 + ((poweredOnVmProvisioned - capacity) / capacity * 100)

      3. Powered Off VMs Datastore:

        datastoreStoragePressurePoweredOff = 100 + ((poweredOffVmProvisioned - capacity) / capacity * 100)

      CPU Overcommitment

      1. All CPU:

        hostCpuPressureAll = 100 + ((totalUsedCores - hostCores) / hostCores * 100)

      2. Powered On VMs CPU:

        hostCpuPressurePoweredOn = 100 + ((totalUsedCoresPoweredOn - hostCores) / hostCores * 100)

      3. Powered Off VMs CPU:

        hostCpuPressureOff = 100 + ((totalUsedCoresPoweredOff - hostCores) / hostCores * 100)

      Aggregation per Cluster

      Once the calculations are performed for each host, they are aggregated at the cluster level using the following method:

      Aggregated Cluster Pressure = (sum of host pressure) / (number of hosts)

      Conclusion

      Understanding these calculations and threshold values is crucial for effective monitoring and management of VM overcommitment. Properly configured monitors help in maintaining optimal performance and avoiding potential resource contention issues within the cluster.